Is there any way to estimate the motion effects on anatomical MRI data?
Since the motion may impact the SNR, whether SNR could act as an index for the motion artifacts?
And, How can I estimate the SNR from one anatomical MRI volume?
people have used the entropy of a single image as a kind of proxy for SNR. However, there are some problems with that (or any other proxy) in your context.
If you have two images acquired in the same subject, one affected by motion and one not, then the one affected by motion will have bigger entropy. BUT, if you have two images acquired in different subjects they will also have different entropy. This potentially
means that subject A with motion might have bigger entropy than subject B without.
